DESCRIPTION (adapted from the Abstract): The HIV epidemic is much better understood now that it was even five years ago, but it still presents daunting problems. Surveillance of the epidemic is one of our most useful endeavors, usually achieved by observing AIDS incidence, not by direct measurement of HIV incidence. Because of the long incubation between HIV infection and AIDS, surveillance of AIDS provides only an indirect measure; hence statistical models play a central role in our understanding of the epidemic. This Principal Investigator proposes to continue to develop analytical techniques needed to extract information from the surveillance system, especially new methods to overcome especially the effects of the long incubation period. Further, he and his collaborators will develop the needed methods to model more effectively the component parts of the epidemic and to further our understanding of how treatment is impacting on the epidemic. He expect that this methodology, in turn, may be used to plan more efficiently and understand more fully the effect of treatment therapies. With the rise of highly effective antiviral treatments has come the development of resistant strains of HIV. Should those resistant strains become widely dispersed, newly diagnosed HIV infection may become even harder to treat. For this reason, the Investigator and his collaborators will study clusters of genotypic patterns associated with resistance among newly diagnosed people with HIV. The observation of such clusters will provide information about how and where such epidemics of resistant virus are emerging. Also, the Investigator proposes to continue to study techniques for making HIV screening test more accurate and cheaper. Thus far, these researchers have proposed and examined various novel way of pooling and retesting blood samples These have achieved, on paper, the twin aims of being more accurate and cheaper. Now with the advent of PCR tests with their advantageous accuracy and direct measurement of the presence of virus (as opposed to the antibody tests) only their high cost seems to block their general applicability/utility. Using the proposed pooling methods will make the PCR tests more widely affordable.